Hello QuNex’s experts,
I have downloaded unprocessed HCP YA dataset on Balsa.
I am preparing QuNex for preprocessing HCP Young Adults raw dataset using the HCP Pipelines. I am focused on rs-fMRI data and would like to reproduce the Minimal preprocessing pipeline without additional denoising ICA steps at the end.
Would anyone already have compiled a “parameters.txt” file that is representative of all parameters used for this dataset with the HCP Pipelines?
Best regards,
Matthieu
Hi Matthieu,
Below you can find the parameters file that I used for HCYA data when I run QuNex acceptance tests before a release. If something seems odd, let me know.
Best, Jure
hcya_parameters.txt (4.6 KB)
Hi Jure,
Thanks for it! I’m going to investigate it.
Wouldn’t you have the gradient coefficients file corresponding to the Siemens Skyra in order to correct for the gradient distortion?
Best,
Matthieu
Hi Jure,
Some questions about the following parameters:
1/ Wouldn’t these Pre FreeSurfer parameters useless since the gradient distortion correction use FILEDMAP and not TOPUP with inversely encoded SE images :
–hcp_sephaseneg : LR
–hcp_sephasepos : RL
–hcp_seechospacing : 0.000580002668012
–hcp_seunwarpdir : z
2/ I would like to get volume outputs of the fMRI minimal preprocessing pipeline without denoising applied. Are the following sections useful?
_image_source : hcp
_image_target : cifti
_bolds : rest
_TR : 0.72
_omit : 0
_pignore : hipass=linear|regress=spline|lopass=linear
_log : keep
_glm_matrix : both
_glm_residuals : save
_bold_nuisance : m,V,WM,WB,1d
_event_file :
_nprocess : 0
_bold_prefix :
_event_string :
_event_file :
_boldname : bold
_shrinknsroi : true
_mov_dvars: 3.0
_mov_dvarsme: 1.6
_mov_fd: 0.5
_mov_radius: 50.0
_mov_scrub: yes
_mov_fidl: udvarsme
_mov_plot: mov_report
_mov_post: udvarsme
_mov_before: 0
_mov_after: 0
_mov_bad: udvarsme
_mov_mreport: movement_report.txt
_mov_preport: movement_report_post.txt
_mov_sreport: movement_scrubbing_report.txt
_mov_pdf: movement_plots
Best,
Matthieu
Hi Matthieu,
Each parameters is only used if the command that uses the parameters gets executed. If you are not running denoising commands, then the parameters for denoising won’t get used. The same holds for se parameters. The test parameters file I provided is not the most clean, so feel free to clean it up for your own needs. I guess this answers both points, just delete the parameters you think are not needed. QuNex will complain quickly if one of the command that you will execute will actually need a parameter.
I always use “cleaned” up parameter files for proper processing to have better overview about what is happening.
Best, Jure
Hi Jure,
Thanks for the tips!
Wouldn’t you have the gradient coefficients file corresponding to the Siemens Skyra in order to correct for the gradient distortion?
Best,
Matthieu
Grad coefficient files are proprietary and I am not allowed to share them with you. 
The HCP release document states this:
Appendix 3. Skyra gradient field nonlinearity coefficients for the HCP Connectome
Skyra. These coefficients are considered proprietary information by Siemens. To
request access to these coefficients, please contact your Siemens collaboration
manager or email Dingxin Wang at dingxin.wang@siemens.com.
Best, Jure
Ok Jure,
I have contacted the appropriate person for my place (For rest of world: Martin Stoltnow) but with no response at the moment
Best,
Matthieu
I guess what you could consider is to instead download data that has been already processed and do just what needs to be done on top. HCP has pretty much all of thair preprocessed data sets uploaded somewhere.
Not sure what you are tyring to do, so maybe this route does not make sense to you. But more often than not, re-processing 1000+ HCYA sessions through the preprocessing pipeline is just a bunch of redundant and time-consuming work.
Best, Jure
Thanks for the advice Jure.
I need volume preprocessed rs-fMRI but not with the denoising (sICA + tICA) part, which potentially removes the effects of physiological signals whose I care about. Unfortunately, on BALSA available preprocessed 3T resting state data only contains the final volume one including denoising…
I couldn’t find anywhere these intermediate preprocessed rs-fMRI volume data. If you have an idea, I take it!
Best,
Matthieu
Hm, never needed such data. I usually neded denoised versions 
. You can try asking at the official HCP forum => https://groups.google.com/a/humanconnectome.org/g/hcp-users.
Best, Jure